Add external links to links.json artifact

docs/source/docset.yml

@@ -74,3 +74,4 @@ toc:
       - file: index.md
       - file: req.md
       - folder: nested
+      - file: external-links.md

docs/source/testing/external-links.md

@@ -0,0 +1,9 @@
+---
+title: "External Links"
+---
+
+[inline external link](kibana://path/to/file.md)
+
+[reference external link][1]
+
+[1]: kibana://path/to/another/file.md

src/Elastic.Markdown/IO/LinkReference.cs

@@ -14,19 +14,24 @@ public record LinkReference
 	[JsonPropertyName("url_path_prefix")]
 	public required string? UrlPathPrefix { get; init; }
 
-	[JsonPropertyName("links")]
-	public required string[] Links { get; init; } = [];
+	[JsonPropertyName("internal_links")]
+	public required string[] InternalLinks { get; init; } = [];
+
+	[JsonPropertyName("external_links")]
+	public required string[] ExternalLinks { get; init; } = [];
 
 	public static LinkReference Create(DocumentationSet set)
 	{
-		var links = set.FlatMappedFiles.Values
-			.OfType<MarkdownFile>()
-			.Select(m => m.RelativePath).ToArray();
+		var markdownFiles = set.FlatMappedFiles.Values.OfType<MarkdownFile>().ToArray();
+		var internalLinks = markdownFiles.Select(m => m.RelativePath).ToArray();
+		var externalLinks = markdownFiles.SelectMany(m => m.Links).ToHashSet().ToArray();
 		return new LinkReference
 		{
 			UrlPathPrefix = set.Context.UrlPathPrefix,
 			Origin = set.Context.Git,
-			Links = links
+			InternalLinks = internalLinks,
+			ExternalLinks = externalLinks
 		};
 	}
+
 }

src/Elastic.Markdown/IO/MarkdownFile.cs

@@ -10,6 +10,7 @@
 using Markdig;
 using Markdig.Extensions.Yaml;
 using Markdig.Syntax;
+using Markdig.Syntax.Inlines;
 using Slugify;
 
 namespace Elastic.Markdown.IO;
@@ -42,6 +43,7 @@ public string? NavigationTitle
 		get => !string.IsNullOrEmpty(_navigationTitle) ? _navigationTitle : Title;
 		private set => _navigationTitle = value;
 	}
+	public IReadOnlySet<string> Links { get; private set; } = new HashSet<string>();
 
 	//indexed by slug
 	private readonly Dictionary<string, PageTocItem> _tableOfContent = new();
@@ -56,6 +58,8 @@ public string? NavigationTitle
 
 	private bool _instructionsParsed;
 
+	public static readonly HashSet<string> ExcludedExternalLinkSchemes = ["http", "https", "mailto", "tel", "file"];
+
 	public MarkdownFile[] YieldParents()
 	{
 		var parents = new List<MarkdownFile>();
@@ -82,11 +86,22 @@ public async Task<MarkdownDocument> ParseFullAsync(Cancel ctx)
 			await MinimalParse(ctx);
 
 		var document = await MarkdownParser.ParseAsync(SourceFile, YamlFrontMatter, ctx);
+		ReadLinks(document);
 		if (Title == RelativePath)
 			Collector.EmitWarning(SourceFile.FullName, "Missing yaml front-matter block defining a title or a level 1 header");
 		return document;
 	}
 
+	private void ReadLinks(MarkdownDocument document)
+	{
+		var links = document.Descendants<LinkInline>()
+			.Select(l => l.Url)
+			.Where(l => !string.IsNullOrEmpty(l) && !l.StartsWith('/'))
+			.Where(url => Uri.TryCreate(url, UriKind.Absolute, out var uri) && !ExcludedExternalLinkSchemes.Contains(uri.Scheme));
+
+		Links = new HashSet<string>(links!);
+	}
+
 	private void ReadDocumentInstructions(MarkdownDocument document)
 	{
 		if (document.FirstOrDefault() is YamlFrontMatterBlock yaml)

src/Elastic.Markdown/Myst/InlineParsers/DiagnosticLinkInlineParser.cs

@@ -44,6 +44,7 @@ public override bool Match(InlineProcessor processor, ref StringSlice slice)
 			return match;
 
 		var url = link.Url;
+		var uri = Uri.TryCreate(url, UriKind.Absolute, out var u) ? u : null;
 		var line = link.Line + 1;
 		var column = link.Column;
 		var length = url?.Length ?? 1;
@@ -58,7 +59,12 @@ public override bool Match(InlineProcessor processor, ref StringSlice slice)
 			return match;
 		}
 
-		if (Uri.TryCreate(url, UriKind.Absolute, out var uri) && uri.Scheme.StartsWith("http"))
+		if (uri != null && !MarkdownFile.ExcludedExternalLinkSchemes.Contains(uri.Scheme))
+		{
+			return match;
+		}
+
+		if (uri != null && uri.Scheme.StartsWith("http"))
 		{
 			var baseDomain = uri.Host == "localhost" ? "localhost" : string.Join('.', uri.Host.Split('.')[^2..]);
 			if (!context.Configuration.ExternalLinkHosts.Contains(baseDomain))
@@ -128,8 +134,5 @@ public override bool Match(InlineProcessor processor, ref StringSlice slice)
 			link.Url += $"#{anchor}";
 
 		return match;
-
-
-
 	}
 }